US customs agency says tariff refund system progressing, but payments may take up to 45 days

Key event: U.S. Customs is building a refund system for roughly $166 billion in tariffs the Supreme Court struck down under IEEPA, with portal development 60–85% complete and phased claim acceptance planned; reviews could take up to 45 days once live. So far 26,664 importers (78% of affected entries) have completed e-refund setup covering about $120 billion; over 330,000 importers and 53 million shipments were affected. The Court of International Trade was left to resolve refund procedures; a judge ordered CBP to use the existing system but CBP proposes a new phased process that would not require importers to sue.

Analysis

An expedited flow of large, retroactive cash refunds to importers would act like a one-time liquidity pulse that disproportionately benefits larger, well-integrated buyers and the banks that service them. Expect working-capital resets (lower payables, higher cash balances) to free up 1-3% of net leverage for large importers within 1-3 months if the portal accepts claims in phases, which could drive near-term buybacks, inventory replenishment and higher freight demand as deferred orders clear.

Logistics providers with scale and claims-handling capabilities stand to capture share from smaller importers that drop out of the reimbursement process; this is a structural redistribution of throughput rather than a pure volume expansion. That benefits asset-light integrators and 3PLs that can bill and collect quickly while exposing full-asset carriers to temporary rate compression as shippers re-negotiate now-corrected landed costs.

Political and legal friction is the largest amplifier: administrative slow-walking, selective prioritization of claim types, or litigated disputes over interest and offsets could stretch real cash delivery from months to years, reversing any short-term bounce. Monitor three binary catalysts across distinct horizons — portal go-live phases (days–weeks), claim-processing velocity and interest/offset rulings (months), and any regulatory attempts to reimpose or replace duties (quarters–years) — as each will reprice sector winners and losers.

Net of these dynamics, markets have underweighted the asymmetric benefits to scale and claims-processing capability versus pure exposure to import volume; the right exposures are therefore to operators that both handle disputes and monetize regained flows rapidly, not to exporters or commodity winners.

Key Decisions for Investors

• Long FDX (3–9 months): buy shares sized 2–3% portfolio or buy a 3–6 month call spread to limit downside. Thesis: capture share from smaller importers and benefit from higher freight velocity as refunds free working capital. Reward: 15–25% upside if processing is timely; Risk: -15% if administrative delays persist or rate competition compresses margins; stop-loss at -12%.
• Long SMCI (12+ months): buy LEAP calls (12–18 months) or 3–5% position in equity for secular AI demand that is insulated from tariff noise. Thesis: hardware demand driven by AI remains the main growth engine independent of tariff tailwinds; refunds are a positive macro crosswind but not the primary driver. Reward: 40–80% upside if execution continues; Risk: supply-chain disruption or cyclical server demand reset could halve gains; keep position size manageable.
• Tactical pair (3–6 months): long FDX / short a small-cap, import-exposed industrial (size net zero). Thesis: scale wins the refund-processing lottery — larger integrators get faster cash and pricing power while smaller importers fail to recapture costs. Reward: asymmetric capture of spread between winners and losers ~10–20%; Risk: policy reversal that re-favors domestic protection or systemic shipping slowdown (-10%).